Google Play Store Listing Conversion

A grocery list app can have a crowded search result and an empty install. Play Console will not describe that gap with Apple’s phrase “unique device impressions.” It has its own counts, and the help page names them without handing you a slogan about what “good” looks like. Read the counts Play defines. Change the listing with the tools Play actually gives you: experiments, and custom listings aimed at a segment. Then read the same counts again.

The names are in View app statistics . That article points at a store listing conversion analysis page for the rate. How the tap before the visit behaves is in Google Play install clicks. Setting up a test is in store listing experiments, not repeated here. Apple’s different fraction is in App Store conversion rate.

The two counts Play actually defines

Store listing visitors: the number of users that visited your store listing who did not have your app installed on any device. Store listing acquisitions: the number of users that visited your store listing and installed your app, who did not have your app installed on any device. Both sentences exclude people who already had the grocery app on some phone or tablet. A reinstall by someone who still has it on a tablet is not this visitor, and it is not this acquisition.

The same help page says the store listing conversion analysis page is where you understand the conversion rate of your store listing. It does not print an equation in that article. Do not promote “acquisitions divided by visitors” as if Google published that sentence, even though the two names sit next to each other. Read the rate on the conversion analysis page. If you divide the two counts yourself, label the result as your ratio and expect it to disagree when filters, dates, or the page’s own rules differ.

What a grocery list visit is

A visitor opened the listing and did not already have the app on any device. They are not “anyone who saw you in search.” Search can show the icon, the name, the short description, and the rating without a listing visit. That earlier tap is a different report. If visitors are flat while you celebrate a ranking tool, you may be getting seen and not opened. The install-clicks post is the place for that earlier step. This post starts when they are on the listing.

Visitors and acquisitions both require that the user did not already have the app on any device. The conversion rate is the one on Play’s analysis page, not a ratio you rename.

Visitors up, acquisitions flat

People are opening the grocery listing and not installing. The listing is the object: icon, short description, screenshots, and the rating they see, which Play weights toward recent ratings. A short description that says “the best lists” and screenshots of an empty phone bezel will lose to a listing that shows milk, bread, and a shared aisle. Change those assets. Do not respond to this pattern by rewriting a keyword list and calling it conversion work.

Experiments are how you test an alternate icon, description, or graphic without guessing which edit did it. The setup, the traffic split, and when to call a winner belong in the experiments post. Use that post when you are ready to run one. Custom listings are the other lever, when the default page is fine for one country or one query and wrong for another.

Custom listings are a segment, not a second app

You can create up to 50 custom store listings. The custom store listing article lists who you can target: churned users (they uninstalled), lapsed users (they have not opened the app in the last 28 days), buyers in several shades, ads traffic, country or region, pre-registration, search keywords, and custom audiences you define. The default listing is what people see in countries you do not target.

For each custom listing you can change the name, icon, descriptions, and graphic assets. Contact details, the privacy policy, and the category stay shared. A grocery app can show a Spanish weekly-shop screenshot set to one country, or a different short description to people who arrive on a specific search keyword, without forking the whole app. That is a conversion lever because those visitors see different pixels. It is not 50 thin duplicates of the same screen with a city name swapped in.

Lapsed and churned targeting will not fill the visitor count this article uses, because those people already had the app. They are a return path, measured in audience and retention, not in “first listing visit from someone with zero installs.” Keep that straight or you will “optimize conversion” with a screenshot aimed at people the visitor metric ignores. Country, search keywords, ads, and pre-registration can still change what a new visitor sees.

Do not import Apple’s fraction

Apple’s conversion rate is downloads and pre-orders divided by unique device impressions, and a pre-order counts once. Play’s listing counts are users, they exclude anyone who already has the app on any device, and the rate sits on a Play page that does not use Apple’s denominator. A grocery app on both stores will show two honest numbers that are not supposed to match. Report them under their own names. A blended “store CVR” in a deck is how you ship the wrong screenshot to the wrong console.

What to leave alone

Listing text and graphics on Play can be updated separately from an app bundle. You do not need a release to fix a misleading first screenshot, which is the opposite of iOS metadata. If a staged rollout’s new build requires the listing to change, Google recommends making that listing change after the rollout reaches 100%, so you are not advertising a screen most people cannot install yet. Until then, read visitors and acquisitions for the listing that is actually live.